[sdiy] VCO tuning philosophy re-visited

[David G. Dixon]

> >Would you recommend the LM331 for use in regular audio VCOs too (i.e.,
> >non-ultrasonic ones)?
> 
> That method has never been popular, I believe because the output is a
> narrow pulse.  There is a mV level Saw available also, which you could try
> amplifying.

Still, it begs the question as to why the LM331 clocking an 8-bit counter to
a DAC isn't seen more often as a basic variation on the sawcore VCO.  Of
course, it would give a staircase, but at 8 bits, the steps would be pretty
darned small, and could be filtered to smoothness pretty easily, or
interpolated using Henry Walmsley's scheme.

Ian, you and I discussed this LM331 business a year or so ago (after I
started a thread about Henry Walmsley's 16-step waveform generator) but then
I forgot all about it, but now I'm all excited again, and I think I'm going
to design a wavetable oscillator around it.  I prefer this method to the
ramp-controlled method that Olivier mentioned, because it should be
glitch-proof -- no worries about the ramp being too small or too large,
particularly with changing frequency, and no need for an ADC.